Organic light-emitting diodes (OLED) generate light by re-combination of electrons and holes, and emit light when a bias is applied between the anode and cathode such that an electrical current passes between them. The brightness of the light is related to the amount of the current. If there is no current, there will be no light emission, so OLED technology is a type of technology capable of absolute blacks and achieving almost “infinite” contrast ratio between pixels when used in display applications.
Several approaches are taught in the prior art for pixel thin film transistor (TFT) circuits to deliver current to an element of a display device, such as for example an organic light-emitting diode (OLED), through a drive transistor. In one example, an input signal, such as a low “SCAN” signal, is employed to switch transistors in the circuit to permit a data voltage, VDAT, to be stored at a storage capacitor during a programming phase. When the SCAN signal is high and VDAT is isolated from the circuit by a switch transistor closing, the VDAT voltage is retained by the capacitor and this voltage is applied to a gate of a drive transistor. With the drive transistor having a threshold voltage VTH, the amount of current to the OLED is related to the voltage on the gate of the drive transistor by:
      I    OLED    =            β      2        ⁢                  (                              V            DAT                    -                      V            OLED                    -                      V            TH                          )            2      (β is a constant related to the properties of the drive transistor).
TFT device characteristics, especially the TFT threshold voltage VTH, may vary, for example due to manufacturing processes or stress and aging of the TFT device during the operation. With the same VDAT voltage, the amount of current delivered by the drive TFT could vary by a large amount due to such threshold voltage variations. Therefore, pixels in a display may not exhibit uniform brightness for a given VDAT value.
Conventionally, therefore, OLED pixel circuits have high tolerance ranges to variations in threshold voltage and/or carrier mobility of the drive transistor by employing circuits that compensate for mismatch in the properties of the drive transistor. For example, an approach is described in U.S. Pat. No. 7,414,599 (Chung et al., issued Aug. 19, 2008), which describes a circuit in which the drive TFT is configured to be a diode-connected device during a programming period, and a data voltage is applied to the source of the drive transistor.
With such circuit configuration, however, the anode of the OLED is not reset in relation to the programming phase. Rather, there will be residual voltage at the OLED anode. When emission starts and the emission current flows through the OLED during the emission phase, the OLED will need some time to refresh the data voltage at the anode. A first problem with this is that it may affect the true black state. If the previous frame data voltage corresponds to a white grayscale and the current frame data voltage corresponds to a black grayscale, for example, there will be some light emission due to the residual voltage at the beginning of the emission phase. The true black state will be compromised. A second problem is memory effects from the previous frame data. If the programmed current is a low current, it could take a significant time to refresh the anode to the programmed value. During the refresh period, the light emission could vary due to the previous residual data at the anode of the OLED, which means the same programmed data could have different light emission as affected by the previous frame data, especially for relative small emission current.
Another approach is described in U.S. Pat. No. 8,314,788 (Choi, issued Nov. 12, 2012). In such circuit, the diode-connection voltage for a drive transistor is pulled down by changing the voltage level at the top plate of the storage capacitor. There are significant drawbacks with such configuration and method. First, when pulling down the gate voltage of the drive transistor, the diode-connected drive transistor is forward biased, and there could be a large instant current to the OLED. This may cause an instance of high luminance light, which would prevent a pixel from ever having a true black state. Second, the anode of the OLED and the gate voltage of the drive transistor would hold the voltage from the previous frame. As there is no initialization or reset scheme, the voltage from the previous frame could affect the programmed voltage for the current frame. Therefore, the current to the OLED during a frame may be affected by the state in the previous frame, as well as by the applied data.
Other approaches to address the above problems have proven deficient. U.S. Pat. No. 7,936,322 (Chung et al., issued May 3, 2011) describes a scheme to reduce the number of transistors to five by overlapping scan and emission control signals. This approach, however, could cause a leakage current during the programming phase, which may affect the blackness in low current operations. U.S. Pat. No. 8,237,637 (Chung, issued Aug. 7, 2012) describes a scheme to improve the blackness and remove the memory effects on the anode of the OLED by adding one more transistor between the initial voltage and the anode. This configuration, however, increases the transistor number to seven in the circuit, which will lower the yield and be difficult to implement in high resolution applications requiring a small geometry. U.S. Pat. No. 8,314,788 (Choi, issued Nov. 12, 2012) describes a scheme to reduce the number of transistors by pulling down the gate voltage of the drive transistor to save the isolation transistor between the driving transistor and OLED. With such configuration, however, there could be high instant current during programming phase. There is no initialization scheme, and thus memory effects could affect the true blackness. U.S. Pat. No. 9,337,439 (Kwon, issued May 10, 2016) describes a scheme to improve the blackness by using previous data, but the number of transistors is still high and the residual voltage at the anode of the OLED could still cause some light leakage for low emission current. U.S. Pat. No. 9,489,894B2 (Yin et al.) describes a scheme to use ELVDD as an initial signal, which reduces the signal lines by one. This approach, however, still has the same number of transistors as U.S. Pat. No. 7,414,599, and the same residual memory effects at the anode of the OLED.